Jurnal Politeknik Negeri Batam (PoliBatam)
Not a member yet
3001 research outputs found
Sort by
Automatic License Plate Detection System with YOLOv11 Algorithm
No full textThe increasing number of motor vehicles in Indonesia demands technological solutions to enhance efficiency and security, particularly in automatic license plate recognition systems. This study aims to develop an automatic license plate detection system using the YOLOv11 algorithm to detect license plates and their characters in real-time. The research methodology includes collecting datasets from Kaggle, RoboFlow, and manual acquisition, followed by annotation, data augmentation, model training, and interface development using Tkinter and OpenCV. The dataset comprises 4000 license plate images and 3000 characters images, divided for training, validation, and testing. Evaluation results demonstrate strong model performance, with precision of 0.891, recall of 0.911, mAP50 of 0.906, and mAP50-95 of 0.631 for license plate detection, and precision of 0.889, recall of 0.912, mAP50 of 0.907, and mAP50-95 of 0.629 for character detection. Real-time testing showed that 12 out of 12 license plates were successfully recognized, influenced by lighting conditions, distance, and plate orientation. This study produced an efficient system for parking security, with potential for further development
The Influence of Knowledge Management and Digital Competence on Employee Performance: Mediating Role of Innovative Behavior
No full textRapid technological changes in the era of Industry 4.0 and 5.0 have made digital knowledge and skills more important in improving the way employees perform their tasks. Earlier research has given mixed results. This shows there is still a lot to learn. Based on the KBV (Knowledge Based-View) theory, this study looks at how knowledge management and digital competence directly and indirectly affect employee performance through innovative work behavior. Data were obtained using a questionnaire that had been compiled and analyzed with Partial Least Squares-Structural Equation Modeling (PLS-SEM) method with SmartPLS 4.1.1.4. The research sample included all employees in the case study (N = 56), with census sampling method. The study found that KM had a significant impact on IWB (p < 0,05), but did not have a significant direct impact on EP (p > 0,05). DC had a significant impact on EP (p < 0,05), but did not have a significant impact on IWB (p > 0,05). IWB played an important role in improving EP and also mediated the relationship between KM and EP. Theoretically, this study adds value to both the KBV theory by explaining how KM boosts performance through indirect ways, and by showing that digital capital plays a limited role in improving performance. Practically, the findings offer actionable implications for HR practitioners in designing performance systems that reward innovative behaviour, thereby motivating employees to utilize knowledge and digital tools more creatively to enhance productivity and service quality in medium enterprises
Binary Classification for Predicting the Investment Trends of The Younger Generation Based on Machine Learning
No full textThis computational study examines investment behavior patterns among a specialized cohort of 115 final year and thesis writing university students, implementing sophisticated feature engineering to transform categorical survey responses into quantifiable financial metrics. The research methodology leverages this unique dataset where respondents\u27 advanced academic standing provides particularly relevant insights into near-term investment decisions. Experimental outcomes reveal distinct algorithmic performance patterns: Random Forest achieved 69.6% accuracy in multi-class classification with weighted averages of 0.662 precision, 0.696 recall, and 0.678 F1-score, while Logistic Regression demonstrated superior binary classification capability with 82.6% accuracy, supported by 0.818 precision, 0.826 recall, and 0.814 F1-score (weighted averages). The hybrid architecture integrating machine learning with business rules achieved peak performance of 85.2% accuracy, successfully balancing predictive power with operational interpretability. These findings underscore how strategically engineered features combined with a carefully selected respondent pool can effectively decode complex financial behaviors, providing financial institutions with actionable frameworks for developing targeted investment solutions for the graduate student demographic while advancing methodological approaches for specialized survey data in fintech applications
Medical Named Entity Recognition from Indonesian Health-News using BiLSTM-CRF with Static and Contextual Embeddings
No full textNamed Entity Recognition (NER) is vital for structuring medical texts by identifying entities such as diseases, symptoms, and drugs. However, research on Indonesian medical NER remain limited due to the lack of annotated corpora and linguistic resources. This scarcity often leads to difficulties in learning meaningful word representations, which are crucial for accurate entity identification. This research aims to compare the effectiveness of static and contextual embeddings in enhancing entity recognition on Indonesian biomedical text. The experimental setup involved utilizing both static (Word2Vec) and contextual (IndoBERT) embeddings in conjunction with neural architectures (BiLSTM) along with Conditional Random Fields (CRF). The BiLSTM architecture was selected for its ability to capture bidirectional dependencies in language sequences. Specifically, four models: Word2Vec-BiLSTM, Word2Vec-BiLSTM-CRF, IndoBERT-BiLSTM, and IndoBERT-BiLSTM-CRF were evaluated to assess the impact of contextual representations and structured decoding. The models were trained on a manually annotated DetikHealth corpus, where specific medical entities such as diseases, symptoms, and drugs were labeled with the BIO-tagging scheme. Performance was subsequently evaluated based on standard metrics: precision, recall, and F1-score. Results indicate that IndoBERT’s contextual embeddings significantly outperform static Word2Vec features. The IndoBERT-BiLSTM-CRF model achieved the highest performance micro-F1 0.4330, macro-F1 0.3297, with the Disease entity reaching an F1-score of 0.5882. Combining contextual embeddings with CRF-based decoding enhances semantic understanding and boundary consistency, demonstrating superior performance for Indonesian biomedical NER. Future work should explore domain-adaptive pretraining and larger biomedical corpora to further improve contextual accuracy
Security Evaluation of Keycloak-Based Role-Based Access Control in Microservice Architectures Using the OWASP ASVS Framework
No full textThe Rocket Car Wash Semarang application operates using a microservice architecture that handles sensitive information such as user identity data, transaction history, and vehicle details. As multiple services interact through authenticated API calls, strong access control is required to protect the system from unauthorized access and privilege escalation. This research evaluates the Keycloak-based Role-Based Access Control (RBAC) implementation by referencing relevant domains of the OWASP Application Security Verification Standard (ASVS) Level 2, specifically V2: Authentication, V3: Session Management, V4: Access Control, and V14: Configuration. The RBAC structure consists of three primary roles—Admin, Owner, and Customer—and the assessment examines the correctness of role–permission mapping and token-based authorization across microservices. The security evaluation was conducted through configuration auditing, API endpoint verification using Postman, JWT token validation, and automated penetration testing using OWASP Zed Attack Proxy (ZAP). The ZAP scan targeted common web vulnerabilities, particularly misconfigurations and weaknesses in HTTP security headers. The results indicate that Keycloak effectively enforces centralized authentication and authorization, with no critical issues such as Broken Access Control identified. However, several non-critical weaknesses were found, including incomplete Content Security Policy (CSP) directives and missing HSTS headers. These findings show that the RBAC implementation meets core ASVS Level 2 controls, while further improvements in security header configuration are required to enhance overall system resilience.The Rocket Car Wash Semarang application operates using a microservice architecture that handles sensitive information such as user identity data, transaction history, and vehicle details. As multiple services interact through authenticated API calls, strong access control is required to protect the system from unauthorized access and privilege escalation. This research evaluates the Keycloak-based Role-Based Access Control (RBAC) implementation by referencing relevant domains of the OWASP Application Security Verification Standard (ASVS) Level 2, specifically V2: Authentication, V3: Session Management, V4: Access Control, and V14: Configuration. The RBAC structure consists of three primary roles—Admin, Owner, and Customer—and the assessment examines the correctness of role–permission mapping and token-based authorization across microservices. The security evaluation was conducted through configuration auditing, API endpoint verification using Postman, JWT token validation, and automated penetration testing using OWASP Zed Attack Proxy (ZAP). The ZAP scan targeted common web vulnerabilities, particularly misconfigurations and weaknesses in HTTP security headers. The results indicate that Keycloak effectively enforces centralized authentication and authorization, with no critical issues such as Broken Access Control identified. However, several non-critical weaknesses were found, including incomplete Content Security Policy (CSP) directives and missing HSTS headers. These findings show that the RBAC implementation meets core ASVS Level 2 controls, while further improvements in security header configuration are required to enhance overall system resilience
Optimization of Spatial Disaster Profile Database for Spatial Disaster Risk Analysis
No full textThis study developed a dynamic, web-based integrated spatial disaster profile database system that is highly vulnerable to various types of natural hazards, using Lebak Regency as a case study. The reference for each profile displayed is the Indonesian Disaster Risk (RBI). The output of this study is a web performance overview consisting of an interactive HTML-based frontend integrated with the backend spatial data management using MongoDB, Python, and JavaScript. This system provides district-level statistical summaries, visualizations with thematic classifications, and an automatic update feature via API simulation. In addition, this system integrates spatial and non-spatial data. Based on the evaluation, this system improves the effectiveness of data collection and utilization, supports evidence-based decision making, and strengthens cross-sector collaboration. The use of a non-relational database architecture optimized for dynamic spatial data with synchronous updates and web-based distribution is a major innovation with the hope of creating a standardized and adaptive disaster information system that can be replicated in other regions with similar risks
Flood Vulnerability Using Multi-Criteria Analysis in West Jakarta
No full textThe Special Capital Region of Jakarta, located on the north coast of Java Island, is a low-lying sedimentary area crossed by 13 rivers, making it highly vulnerable to flooding from river overflows and sea level rise intensified by climate change. Flooding has been a persistent issue since the Dutch colonial era, driven by rapid urban growth, inadequate development management, the emergence of slums, and poor water resource management. Major floods occurred in 2002 and 2007, and annual floods continued until 2020. Several key factors contribute to flooding, including slope, rainfall, and land cover, while in West Jakarta the problem is further exacerbated by land subsidence and sea level rise. To address this, a flood vulnerability analysis was carried out using a GIS-based multi-criteria analysis approach. Unlike the BNPB reference, which considers only slope, rainfall, and land cover, this study incorporated six parameters: land cover, rainfall, drainage density, slope, land subsidence, and sea level rise. Each parameter was weighted based on its contribution to flooding risk, and spatial data were processed through overlay and reclassification to produce a flood vulnerability map. The results categorize West Jakarta into four classes: Not Vulnerable, Slightly Vulnerable, Vulnerable, and Highly Vulnerable. The majority of the area falls into Not Vulnerable (39.62%) and Slightly Vulnerable (36.30%), while only Kembangan District is dominated by the Highly Vulnerable category. Validation was conducted by comparing the vulnerability map with historical flood data from 2016 to 2023. The comparison shows a strong positive correlation between mapped vulnerability levels and the frequency of actual flood events. This alignment confirms the validity of the method and highlights its value for improving flood risk management and mitigation planning in West Jakarta
The Influence of Artificial Intelligence on Student’s Intention to Commit Academic Fraud with Family Culture as a Moderating Variable
No full textThe purpose of this study is to examine the influence of artificial intelligence on students\u27 intention to commit academic fraud, with family culture moderating the effect of artificial intelligence on this intention. Data were collected using questionnaires distributed both online and offline, with a sample of 120 accounting students from four universities in Batam City. The analysis used descriptive statistics. The results of the study indicate that artificial intelligence has a positive influence on students\u27 intention to commit academic fraud, and that family culture strengthens the effect of artificial intelligence on this intention
Sosialisasi PPN atas Perdagangan E-Commerce di Kota Pontianak
No full textThe Community Service Program (PKM) themed “Tax Update: Socialization of Value Added Tax (VAT) on E-Commerce Transactions in Pontianak City” was conducted to improve tax literacy and raise public awareness of digital VAT obligations. The background of this activity is the rapid growth of e-commerce and the limited understanding of VAT regulations. The methods included delivering presentations, interactive discussions, and evaluation through questionnaires completed by 66 participants. The evaluation focused on four variables: tax understanding, tax socialization, taxpayer compliance, and VAT revenue. The results showed that participants’ tax understanding was categorized as Good (4.32), tax socialization as Very Good (4.52), taxpayer compliance as Good (4.37), and VAT revenue as Very Good (4.54).In conclusion, this PKM activity effectively enhanced the knowledge and tax awareness of the people of Pontianak, while also encouraging stronger commitment to contributing to state revenue through compliance with e-commerce VAT.Program Pengabdian kepada Masyarakat (PKM) bertema “Tax Update: Sosialisasi Pajak Pertambahan Nilai (PPN) atas Transaksi E-Commerce di Kota Pontianak” dilaksanakan untuk meningkatkan literasi pajak dan kesadaran masyarakat terhadap kewajiban PPN digital. Latar belakang kegiatan ini adalah semakin berkembangnya e-commerce serta masih terbatasnya pemahaman terkait regulasi PPN.Metode kegiatan mencakup penyampaian materi, diskusi interaktif, dan evaluasi melalui kuesioner yang diikuti oleh 66 peserta. Evaluasi difokuskan pada empat variabel: pemahaman pajak, sosialisasi pajak, kepatuhan wajib pajak, dan penerimaan PPN. Hasil menunjukkan bahwa pemahaman pajak peserta berada pada kategori Baik (4,32), sosialisasi pajak Sangat Baik (4,52), kepatuhan wajib pajak Baik (4,37), dan penerimaan PPN Sangat Baik (4,54).Kesimpulannya, kegiatan PKM ini efektif meningkatkan pengetahuan dan kesadaran pajak masyarakat Pontianak, serta mendorong komitmen untuk berkontribusi terhadap penerimaan negara melalui kepatuhan PPN e-commerce
Pembuatan Website sebagai Media Informasi Desa Wisata Nagari Sungai Batang
No full textNagari Sungai Batang, known as the birthplace of the religious figure Buya Hamka, is a culturally rich village with great tourism potential. However, it currently lacks an official website, and information about the village is only available on external news portals or the Ministry of Tourism and Creative Economy website. To enhance its visibility and accessibility, this community service project aimed to develop a dedicated website for Nagari Sungai Batang as a medium of information and promotion. This activity was part of a collaborative program between Politeknik Negeri Batam and Politeknik Negeri Padang, where the former developed the website while the latter created the information system. The website was developed using the Waterfall model, covering stages from requirements analysis to implementation and testing. Evaluation was conducted using Alpha-Beta testing, involving both developer and user environments. After completion, the team delivered the website and conducted user training on content management. The result showed that the website effectively served as an information and promotional medium, achieving a user satisfaction rate of 96%.Nagari Sungai Batang merupakan desa kelahiran tokoh agama Buya Hamka yang memiliki potensi wisata dan nilai budaya tinggi. Namun, desa ini belum memiliki website resmi, sehingga informasi mengenai potensi wisata masih tersebar di situs berita atau di laman Kementerian Pariwisata dan Ekonomi Kreatif. Agar dapat dikenal lebih luas, dibutuhkan media informasi yang dapat dikelola langsung oleh pihak pengelola desa. Kegiatan pengabdian ini bertujuan untuk membuat website sebagai sarana promosi dan penyebaran informasi Desa Wisata Nagari Sungai Batang. Kegiatan ini merupakan bagian dari kolaborasi antara Politeknik Negeri Batam dan Politeknik Negeri Padang, di mana Politeknik Negeri Batam berperan dalam pengembangan website, sedangkan Politeknik Negeri Padang mengembangkan sistem informasinya. Tahapan kegiatan meliputi identifikasi kebutuhan, pengembangan, pengujian, serta penyerahan dan sosialisasi penggunaan website kepada pengelola desa, khususnya dalam pengelolaan konten. Hasil evaluasi menunjukkan bahwa website yang dikembangkan efektif sebagai media informasi dan promosi, dengan tingkat kepuasan pengguna mencapai 96%